Create estimation templates from MixAR model objects

Create estimation templates from MixAR model objects.

Usage

est_templ(model, shift = TRUE, ...)

Arguments

model

a "MixAR" object.

shift

logical, see Details.

...

currently not used.

Details

Argument model is used as a template to specify values of parameters and/or which parameters to estimate or fix. In general, If a value of a parameter in model is NA, then it is to be estimated. Otherwise the parameter is taken as is.

The current implementation is incomplete. In particular, the AR parameters are always designated for estimation.

Argument shift can be used to overwrite some or values component shift in model. If shift has length one, it is replicated to the number of MixAR components. If shift[k] is TRUE, then the shift coefficient for the k-th component is set to NA to request its estimation. Otherwise, the value of the shift for the k-th component in model is taken.

Argument shift has a default of TRUE which causes the shift coefficients to be estimated irrespectively of their values in model.

est_templ returns a list with as many components as there are MixAR components in the model. The k-th component of the list is itself a list specifing which parameters of the i-th MixAR component to estimate or fix.

Value

a list, as described in Details.

Examples

exampleModels$WL_A
#> An object of class "MixARGaussian"
#> Number of components: 2 
#>        prob shift scale order ar_1
#> Comp_1 0.5    0     5     1   0.5 
#> Comp_2 0.5    0     1     1   1.1 
#> 
#> Distributions of the error components:
#> 	standard Gaussian
#> 
est_templ(exampleModels$WL_A)
#> [[1]]
#> [[1]][[1]]
#> [1] NA
#> 
#> [[1]][[2]]
#> [1] NA
#> 
#> 
#> [[2]]
#> [[2]][[1]]
#> [1] NA
#> 
#> [[2]][[2]]
#> [1] NA
#> 
#> 
est_templ(exampleModels$WL_A, shift = FALSE)
#> [[1]]
#> [[1]][[1]]
#> [1] 0
#> 
#> [[1]][[2]]
#> [1] NA
#> 
#> 
#> [[2]]
#> [[2]][[1]]
#> [1] 0
#> 
#> [[2]][[2]]
#> [1] NA
#> 
#> 

exampleModels$WL_I
#> An object of class "MixARGaussian"
#> Number of components: 3 
#>        prob shift scale order ar_1 ar_2  ar_3
#> Comp_1 0.4    0     1     2    0.9 -0.60     
#> Comp_2 0.3    0     1     1   -0.5           
#> Comp_3 0.3   -5     5     3    1.5 -0.74 0.12
#> 
#> Distributions of the error components:
#> 	standard Gaussian
#> 
est_templ(exampleModels$WL_I)
#> [[1]]
#> [[1]][[1]]
#> [1] NA
#> 
#> [[1]][[2]]
#> [1] NA NA
#> 
#> 
#> [[2]]
#> [[2]][[1]]
#> [1] NA
#> 
#> [[2]][[2]]
#> [1] NA
#> 
#> 
#> [[3]]
#> [[3]][[1]]
#> [1] NA
#> 
#> [[3]][[2]]
#> [1] NA NA NA
#> 
#>